Child–robot interaction: designing robot mediation to facilitate friendship behaviors

Figures & data

Figure 1. Robot Skusie.

Figure 2. Two types of robot-mediated interaction activities: conversational and tablet assisted.

Table 1. Friendship behaviors and indicators.

Friendship Behaviors	Behavioral Indicators	Examples of Presence (coded “1”)
Liking	Expressing positive emotions about the interaction activities verbally and non-verbally.	• One child says “I like this.”
		• Another child makes a “ooh” sound with a smile.
		• Absence of such behaviors is coded as “0”.
Togetherness	Helping by sharing information (e.g. answers, hints) to assist the partner and seeking a third-party assistance for the partner; making bodily gestures and movements to assist the partner (e.g. pointing, whispering, eye gaze, touch, hand holding); and, using words like “we” or “us,” which demonstrates a sense of belong or one-ness.	• Xav asks the human facilitator to grab Natalie's hand and help her complete the tablet task.
		• When the microphone falls off from Mat's shirt, Cla tries to put it back on.
Parity	Seeking opinions and recognizing autonomy of the partner, yielding turns to the partner.	• When Robot asks the children to help select a birthday present, Cla asks Mat “Do you want this one or this one?” When Mat points at the car, Cla confirms by saying “The car?”
		• Zac whispered Eli’s name and smiled when Skusie asked them to speak one at a time.
Agreement	Verbal expressions of agreement and acknowledgment, mimicking/repeating the partner’s behaviors, validating what the partner has said.	• Zac mimics Eli’s behavior, both covering their mouths and laughing.
		• Kim claps for Xav when he gets the correct answer.
		• In a Tablet session, Mat places something on the birthday cake they are designing on the tablet, Cla whispers, “Yeah, that's a good idea. It's a great idea.”
Co-construction	Working together to create artifacts, such as making objects, sharing personal experiences, and building upon or elaborating on the partner’s speech.	• Both children are actively engaged in making digital artifacts on the shared tablet.
		• Nat says that she has a dog, Est looks at her, saying “I have a dog too.”

Figure 3. Time series plot of friendship behaviors.

Table 2. Means and standard deviations of friendship behavior variables.

Variables (Friendship behaviors)	Robot-mediated interactions
	Conversational		Tablet assisted
	M	SD	M	SD
Liking	4.54	3.04	2.11	2.63
Togetherness	3.43	2.66	2.60	2.08
Parity	4.86	3.00	6.82	3.90
Agreement	1.27	1.96	0.57	1.08
Co-construction	0.29	0.93	3.43	3.39

Table 3. Wilcoxon signed-rank test results for friendship behavior, not liking behavior.

Friendship behavior	Negative ranks			Positive ranks			Test statistics
	n	Mean rank	Sum of ranks	n	Mean rank	Sum of ranks	Ties	Z	P
Liking	82	52.09	4271.00	16	36.25	580.00	7	−6.559^a	<0.001***
Togetherness	50	51.45	2572.50	37	33.93	1255.50	18	−2.818^a	0.005**
Parity	31	34.11	1057.50	60	52.14	3128.50	14	−4.108^b	<0.001***
Agreement	35	37.97	1329.00	25	20.04	501.00	45	−3.076^a	0.002**
Co-construction	5	16.30	81.50	69	39.04	2693.50	31	−7.050^b	<0.001***

Notes: Negative and positive ranks are differences in friendship behaviors between tablet assisted and conversational conditions.

^a Based on positive ranks.

^b Based on negative ranks.

*** p < 0.001; ** p < 0.01; * p < 0.05.

Figure 4. Box plots of friendship behaviors.